基于GCV的LS_SVM模型选择在个人信用评估中的应用_李娴.pdf

DOI :10.15991/j.cnki.411100.2011.03.001 41 3 () Vol.41　No.3 2011 5 Journal of Henan University (Natural Science) M ay 2011 GCV LS-SVM 李娴1, 2 (1.河南大学 数学与信息科学学院, 开封475001;　2.上海财经大学 统计与管理学院, 上海 200433) 　:, (LS- SVM)GCV New ton-Raphson . Fisher 、Logistic . , . :LS-SVM;GCV;New ton-Raphson ;; :TP391　　　:A :1003-4978(2011)03-0240-06 Research and Application of LS-SVM Model Selection Based on GCV in Individual Credit Appraisal LI Xian1, 2 (1.S chool of Mathematics and I nf ormation S cience, Henan University, K aif eng 475001, China ; 2.School of S tatistics and Management, Shanghai University of Finance and Economics, Shanghai 200433, China) Abstract :In view of massive individual credit data as well as non-linear relation between credit and its influencing factors, a new individual credit risk forecast model is established using LS-SVM, in w hich regularization parameter is selected based on the GCV criterion and the New ton-Raphson algorithm.Distinction effects are compared among the new model, Fisher linear discriminant analysis, Logistic regression as well as generalized additive model.The results show that the proposed method not only has the effective model selection ability, but also has the superior distinction predictive ability. Key words :LS-SVM ;GCV ;New ton-Raphson iteration ;model selection ;individual credit evaluation 0　 , , 、、 .: ;.(ANN ) , , 、、 ., , , ., [ 1] , (GAM )., (SVM ), .Shin, Lee Kim [2] SVM , ANN , SVM A NN .LS- SVM SVM , SVM , “”“” .LS-SVM (), , 　:2010-12-10 　:211 (803) 　:(1969-), , , .:. :GCV LS-SVM 　　241 ., .[3]LS-SVM ,